Run an extended treatment-effects analysis (ATE / ATT / ATC + CATE) — morie_run_treatment_effects

R port of investigation.run_treatment_effects_analysis. Returns a list with:

treatment_effects_summary – data.frame of ATE / ATT / ATC point estimates (SE / CI columns present but NA, matching Python).
cate_subgroup_estimates – data.frame of within-subgroup Hajek IPW CATEs with sandwich-style SEs and Wald 95\% CIs.
analysis_frame – the trimmed data.frame with attached propensity score and IPW weight columns (ps, w_ate, w_att, w_atc).
Convenience scalars ate / att / atc / se / ci_lower / ci_upper / n / method preserved from the previous R surface for backward compatibility.

Usage

morie_run_treatment_effects_analysis(data, treatment, outcome, covariates)

Arguments

data: data.frame containing treatment, outcome, and covariates.
treatment: Treatment column name.
outcome: Outcome column name.
covariates: Character vector of covariate column names.

Value

Named list as described above.

Examples

set.seed(1)
df <- data.frame(d = rbinom(80, 1, 0.4), x1 = rnorm(80), x2 = rnorm(80))
df$y <- df$d + df$x1 + rnorm(80)
res <- try(morie_run_treatment_effects_analysis(df, "d", "y",
                                                c("x1", "x2")))
if (!inherits(res, "try-error")) str(res, max.level = 1)
#> List of 11
#>  $ treatment_effects_summary:'data.frame':	3 obs. of  6 variables:
#>  $ cate_subgroup_estimates  :'data.frame':	0 obs. of  8 variables:
#>  $ analysis_frame           :'data.frame':	80 obs. of  8 variables:
#>  $ ate                      : num 1.13
#>  $ att                      : num 1.11
#>  $ atc                      : num 1.15
#>  $ se                       : num 0.316
#>  $ ci_lower                 : num 0.511
#>  $ ci_upper                 : num 1.75
#>  $ n                        : int 80
#>  $ method                   : chr "IPW-Hajek (ATE/ATT/ATC + CATE)"